Re: An update on [Proposal: described vs stated triple terms]

Peter,

[a table *involving predicates from the rdfs namespace*]
Thanks, that is indeed the query result that I’m aiming for.

These two statements can only both be true if semantic effects from RDFS properties are part of simple entailment or are visible in SPARQL BGP matching with no entailment regime in force.

Of course, I assume RDFS entailment regime in my SPARQL example. In don’t see the issue here.

I am becoming more and more confused, not about the details of this line of discussion, but the entire basis of it.

The discussion, if you want to see it, is about whether to accept or not the extension [2] (spoiler: I am not very keen).

At this point I think the working group needs to concentrate on turning the baseline, which was agreed on, into the documents that the working group needs to produce instead of yet again getting into unfocused discussion on different proposals.  My opinion is that the working group has bent over backwards to accommodate what appear to be minority views to the detriment of making progress on the activities that are mandated by its charter.  This is the point at which proposals should be put to cloture votes so that progress can be made on the required activities of the working group.

I agree, and that’s why I wrote about few important and concrete open issues to discuss or close and move on, related to the baseline, which I summarise again here:


  1.  To finalise about the baseline [1]:
     *   Somebody wants to strengthen the unrestricted abstract syntax, to allow triple terms only in object position.
     *   Somebody wants to weaken the well-formed abstract syntax, just to allow triple terms only in object position.
     *   Somebody does not want to have the unrestricted syntax, but only the well-formed syntax.
     *   Somebody argues that the denotation of triple terms should not be injective.
  2.  To reject once forever:
     *   opacity of IRIs (discussed at length as going against the principles of semantic web IRIs);
     *   singleton property semantics (which breaks the way properties are used in common practice).

In parallel, there is the discussion about extending the baseline:

     *   Somebody wants to add a special rdf:states predicate to deal with the case of triple terms whose structure is in triples in the graph [2]; even if this is accepted, there are several open issues:
        *   (?) this extension should extend RDFS and not RDF;
        *   (?) rdf:states rdfs:subPropertyOf rdf:reifies. should hold in RDFS;
        *   (?) the mapping between RDF-star and LPGs should use rdf:states instead rdf:reifies;
        *   (?) the annotation syntax in Turtle could be mapped into rdf:states instead of rdf:reifies.

—e.

[1] https://github.com/w3c/rdf-star-wg/wiki/RDF-star-%22working-baseline%22

[2] https://github.com/w3c/rdf-star-wg/wiki/Extending-the-baseline-with-%22asserted%22-stuff